Heat Recovery Potential: A Simple Hot-Stream and Cold-Stream Matching Example

Every heat-using plant contains two families of process streams: hot streams that must be cooled down and cold streams that must be heated up. The heat recovery potential of the plant is the heat that can be transferred from the first family to the second inside the process—in heat exchangers—instead of being bought from a hot utility (steam, a fired heater) and dumped to a cold utility (cooling water, air). Pinch analysis is the standard methodology for estimating this potential before any exchanger is designed: by optimizing energy recovery, it minimizes how much energy the utilities must supply. A pinch occurs at the closest point of approach between the hot and cold sides of the network, and designing around that point minimizes utility use.

This article explains the quantities involved and then works a small, deliberately simple hypothetical example end to end: four streams, one chosen minimum approach temperature, a numerical heat cascade, and the resulting utility targets and stream matches.

Hot streams, cold streams, and heat duty

A hot stream releases heat as it cools from its supply temperature to its target temperature; a cold stream absorbs heat as it heats from its own supply to its target. Each stream is characterized by its heat-capacity flow rate, CP—the mass flow rate multiplied by the specific heat, in kW/K—so the heat it releases or absorbs (its duty) is simply CP times its temperature change.

Heating that process streams cannot supply each other comes from a hot utility; cooling that cannot be absorbed internally goes to a cold utility. Minimizing these two utilities is the object of the exercise. (The same matching logic extends to mass exchanger networks, where streams exchange contaminant rather than heat, but this article stays with heat.)

What limits heat recovery between two streams

Two independent constraints cap how much heat one stream can give another. First, the overall balance: a hot stream cannot supply more heat than it releases, and a cold stream cannot absorb more than it needs. Second, the temperature driving force: heat flows only from hotter to colder, and a real exchanger needs a finite temperature difference everywhere. The minimum approach temperature, ΔTmin, is the smallest hot-minus-cold difference the designer will accept; it is an economic choice, because a smaller value permits more recovery but demands more heat-transfer area.

Consider a single pair with illustrative numbers. A hot stream cooling from 140 °C to 40 °C at CP = 2 kW/K releases 200 kW, while a cold stream heating from 20 °C to 130 °C at the same CP requires 220 kW. No matching can beat the balance: at most 200 kW is recoverable regardless of temperatures. With ΔTmin = 10 K, that full amount is achievable—in a countercurrent exchanger (the two fluids flowing in opposite directions, which keeps the driving force as even as possible) the temperature difference is exactly 10 K at both ends (140 vs 130 °C, and 40 vs 30 °C) and constant in between—provided the coldest 10 degrees of the cold stream (20 → 30 °C) is heated by steam instead. Notice that the sensible arrangement places the purchased heat at the cold end; buying steam for the hot end instead would strand most of the hot stream’s heat and multiply the steam bill. Pinch analysis generalizes exactly this reasoning to networks of many streams.

A four-stream example, step by step

Single pairs are easy; real plants match several hot and several cold streams at once. Two hot streams, two cold streams, one hot utility, and one cold utility is precisely the configuration used in published benchmark problem sets for small heat exchanger networks, so it is a natural size for learning the method. The stream data below are hypothetical and chosen for clean arithmetic.

Assume ΔTmin = 10 K throughout.

Step 1: duties and a first-pass balance

The hot streams release 200 + 100 = 300 kW in total; the cold streams demand 140 + 180 = 320 kW. If temperature did not matter, the heating deficit would be only 20 kW. The question is what the temperature constraints do to that number.

Step 2: shift every stream’s temperatures by ΔTmin/2

The problem-table trick is to shift each hot stream’s temperatures down by 5 K and each cold stream’s up by 5 K (half of ΔTmin on each side). On this shifted scale, two streams at the same shifted temperature are exactly ΔTmin = 10 K apart on the real scale, so heat can be cascaded down a single temperature axis without re-checking driving forces. The shifted intervals are: H1, 145 → 45; H2, 85 → 35; C1, 105 → 35; C2, 125 → 65.

Step 3: build the problem table

Between successive shifted temperatures, a stream contributes its CP whenever it spans that interval. The net heat of an interval equals (sum of hot CPs minus sum of cold CPs) multiplied by the interval width: positive means surplus heat to pass downward, negative means a deficit to be filled from above.

The graphical alternative, composite curves—combined hot and cold duty curves plotted against temperature with their closest approach marking the pinch—gives the same information; the problem table is simply its numerical counterpart.

Step 4: the heat cascade gives the targets

Cascading the interval heats from the top gives running totals of +40, +20, −40, −60, −20, −20. The most negative value, −60 kW at the shifted temperature 65 °C, is the minimum hot-utility duty that must be injected at the top: QH,min = 60 kW. Adding 60 kW makes every level non-negative and leaves the values 60, 100, 80, 20, 0, 40, 40 from top to bottom. The zero at shifted 65 °C is the pinch—in real temperatures, hot side 70 °C, cold side 60 °C, where the streams are exactly ΔTmin apart. The cascade leaves 40 kW at the bottom: QC,min = 40 kW of cold utility. The balance closes: 60 + 300 = 320 + 40 = 360 kW.

Step 5: match streams above and below the pinch

For maximum energy recovery (MER), the rules are clear. Energy should not be transferred across the pinch, and the network splits into a hot region above it and a cold region below it. Above the pinch, streams are matched “so that heat is transferred from the coolest possible source”; below the pinch, the opposite. Utilities also have their places: hot utility above the pinch, cold utility below it.

One consistent MER set of matches for the example is:

• Above the pinch: H2 (90 → 70 °C, 40 kW) heats C1 from 60 → 80 °C, with 10 K at both exchanger ends. H1’s first stage (150 → 130 °C, 40 kW) finishes C1 (80 → 100 °C) at a comfortable 50 K at both ends. H1’s remaining above-pinch heat (130 → 70 °C, 120 kW) heats C2 from 60 → 100 °C, where the approach narrows from 30 K to exactly 10 K but never violates ΔTmin. Steam supplies the last 60 kW, raising C2 from 100 → 120 °C.
• Below the pinch: H2 (70 → 40 °C, 60 kW) heats C1 from 30 → 60 °C, again 10 K at both ends. Cooling water removes the last 40 kW from H1 (70 → 50 °C).

Every duty closes: H2 delivers 40 + 60 = 100 kW, H1 delivers 40 + 120 + 40 = 200 kW, C1 receives 140 kW, and C2 receives 180 kW. Notice that no heat crosses the pinch: the exchangers above it involve only portions of streams hotter than 70/60 °C, and those below only colder portions.

Reading the results

The targets say that 260 kW of the 320 kW heating demand (about 81%) can be met by process heat, for a steam bill of just 60 kW, and that only 40 kW of the hot streams’ 300 kW must be rejected to cooling water. Compare this with the naive 20 kW deficit from Step 1: the temperature structure costs 40 kW of theoretical recovery. That is exactly the below-pinch surplus—100 kW of hot heat against 60 kW of cold demand below 70/60 °C—which cannot be lifted across the pinch without breaking the MER rules.

This is the practical meaning of heat recovery potential: not the raw imbalance between hot and cold duties, but the balanced, temperature-feasible transfer that a well-designed network can achieve. Note also that “maximum energy recovery” does not mean zero utilities—some hot and cold utility generally remains, as the 60 kW and 40 kW here show. Finally, many different MER networks can exist for the same targets; the list above is one consistent set, and detailed synthesis (exchanger sequencing, stream splitting, capital-cost trade-offs) is a design task of its own.

Assumptions and limitations

The example assumes steady operation, constant CP (straight-line temperature-heat curves), and fixed supply and target temperatures. Streams that change phase have curved composite segments and need more care. The targets also depend on ΔTmin: choosing a smaller value would raise the recovery potential at the price of more exchanger area, so ΔTmin selection is an economic optimization in itself. Targets are ceilings for energy, not blueprints—an installed network typically recovers slightly less, and cost-optimal designs deliberately relax the MER condition.

Data quality matters as much as the method. In retrofit studies, correctly identifying which streams are fixed and which are flexible across operating modes determines whether calculated savings materialize, a point examined in a thesis on stream identification in pinch analysis that built fixed- and flexible-flow pinch models by hand and validated them with mathematical software.

Key takeaways

• Heat recovery potential comes from two checks: the overall heat balance and the ΔTmin driving force everywhere in the network.
• Shifting temperatures by ΔTmin/2 and cascading heat (the problem-table method) yields the minimum hot and cold utilities and the pinch location.
• Above the pinch, no heat crosses it, and matching follows region-specific rules; utilities sit above and below the pinch respectively.
• In the four-stream example, temperature feasibility cut recoverable heat from a naive 300 kW to 260 kW of the 320 kW demand, fixing targets of 60 kW steam and 40 kW cooling.
• Targets are calculated before design and serve as the benchmark any proposed heat exchanger network should approach.